Glossary

Knowledge graph

A plain-English definition, how it works, an example, and how it is used in enterprise AI.

{ Definition }

What is a knowledge graph?

A network of things and the relationships between them, stored so that both people and software can follow the connections.

How it works

A knowledge graph stores information as nodes (things, such as a supplier or a shipment) and edges (relationships, such as "supplies" or "is late for"). Because the relationships are stored directly, questions that cross many links are easy to answer: which customers are affected by a late shipment from one supplier?

In an enterprise, the knowledge graph is usually built from an ontology. The ontology defines which kinds of things and relationships exist; the graph holds the actual instances and connections, drawn from the systems the business already runs.

Knowledge graphs help AI in two ways. They let it find related facts that would be scattered across many tables, and they let it show its working: the path through the graph that led to an answer.

An example

Illustrative example

In an illustrative supply chain graph, a disrupted shipping lane links to the shipments on it, each shipment links to the orders it fulfils, and each order links to a customer. Following those links answers which customers are at risk in one step.

How NexusAI uses it

NexusAI shows its knowledge graph in several views: Network, Map (grouped by domain, such as procurement, finance and inventory), Lanes, Lineage and Clusters. On the practice enterprise it holds 68 entities, 74 relationships, 15 domains and 115 cross-domain links.

Figures from the practice enterprise come from simulated data, not a customer.

{ FAQ }

Questions about Knowledge graph

What is the difference between a knowledge graph and an ontology?

The ontology is the design: which kinds of things, links, rules and actions exist. The knowledge graph is the content: the actual things and connections, filled in from your systems.

Is a knowledge graph a kind of database?

It can be stored in a graph database, but the idea is broader: information organised as connected things. It can also be built virtually over existing systems.

Get started

Build the intelligent enterprise

Turn enterprise systems into an intelligent operating layer that understands your business and executes operations autonomously.

Live in production within 90 daysNo systems replacedCloud, on-premise or hybrid